Deconstructive angel

The Deconstructive AngelAuthor(s): M. H. AbramsSource: Critical Inquiry, Vol. 3, No. 3 (Spring, 1977), pp. 425-438Published by: The University of Chicago PressStable URL: http://www.jstor.org/stable/1342932Accessed: 10/04/2009 02:51

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=ucpress.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform thatpromotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to CriticalInquiry.

http://www.jstor.org

http://www.jstor.org/stable/1342932?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/action/showPublisher?publisherCode=ucpress

The Tlimits of Pluralism

II

The Deconstructive Angel

M. H. Abrams

Demogorgon. -If the Abysm Could vomit forth its secrets:-but a voice Is wanting ...

-Shelley, Prometheus Unbound

We have been instructed these days to be wary of words like

"origin," "center," and "end," but I will venture to say that this session had its origin in the dialogue between Wayne Booth and myself which centered on the rationale of the historical procedures in my book, Natural Supernaturalism. Hillis Miller had, in all innocence, written a review of that book; he was cited and answered by Booth, then re-cited and re-answered by me, and so was sucked into the vortex of our ex-

change to make it now a dialogue of three. And given the demonstrated skill of our chairman in fomenting debates, who can predict how many others will be drawn into the vortex before it comes to an end?

I shall take this occasion to explore the crucial issue that was raised by Hillis Miller in his challenging review. I agreed with Wayne Booth that pluralism-the bringing to bear on a subject of diverse points of view, with diverse results-is not only valid, but necessary to our understanding of literary and cultural history: in such pursuits the con- vergence of diverse points of view is the only way to achieve a vision in

depth. I also said, however, that Miller's radical statement, in his review, of the principles of what he calls deconstructive interpretation goes beyond the limits of pluralism, by making impossible anything that we would account as literary and cultural history.1 The issue would hardly

1. "Rationality and Imagination in Cultural History: A Reply to Wayne Booth," Criti- cal Inquiry 2 (Spring 1976): 456-60.

425


be worth pursuing on this public platform if it were only a question of the soundness of the historical claims in a single book. But Miller con- sidered Natural Supernaturalism as an example "in the grand tradition of modern humanistic scholarship, the tradition of Curtius, Auerbach, Lovejoy, C. S. Lewis,"2 and he made it clear that what is at stake is the validity of the premises and procedures of the entire body of traditional

inquiries in the human sciences. And that is patently a matter important enough to warrant our discussion.

Let me put as curtly as I can the essential, though usually implicit, premises that I share with traditional historians of Western culture, which Miller puts in question and undertakes to subvert:

1. The basic materials of history are written texts; and the authors who wrote these texts (with some off-center exceptions) exploited the pos- sibilities and norms of their inherited language to say something determinate, and assumed that competent readers, insofar as these shared their own linguistic skills, would be able to understand what

they said. 2. The historian is indeed for the most part able to interpret not only

what the passages that he cites might mean now, but also what their writers meant when they wrote them. Typically, the historian puts his

interpretation in language which is partly his author's and partly his own; if it is sound, this interpretation approximates, closely enough for the purpose at hand, what the author meant.

3. The historian presents his interpretation to the public in the expecta- tion that the expert reader's interpretation of a passage will approximate his own and so confirm the "objectivity" of his interpretation. The worldly-wise author expects that some of his interpretations will turn out to be mistaken, but such errors, if limited in scope, will not

seriously affect the soundness of his overall history. If, however, the bulk of his interpretations are misreadings, his book is not to be accounted a history but an historical fiction.

Notice that I am speaking here of linguistic interpretation, not of what is confusingly called "historical interpretation"-that is, the

categories, topics, and conceptual and explanatory patterns that the historian brings to his investigation of texts, which serve to shape the story within which passages of texts, with their linguistic meanings, serve as instances and evidence. The differences among these organizing categories, topics, and patterns effect the diversity in the stories that different historians tell, and which a pluralist theory finds acceptable. Undeniably, the linguistic meanings of the passages cited are in some

degree responsive to differences in the perspective that a historian

2. "Tradition and Difference," Diacritics 2 (Winter 1972): 6.

426 M. H. Abrams

Spring 1977 427

brings to bear on them; but the linguistic meanings are also in consider- able degree recalcitrant to alterations in perspective, and the historian's

fidelity to these meanings, without his manipulating and twisting them to fit his preconceptions, serves as a prime criterion of the soundness of the

story that he undertakes to tell. One other preliminary matter: I don't claim that my interpretation

of the passages I cite exhausts everything that these passages mean. In his review, Hillis Miller says that "a literary or philosophical text, for Abrams, has a single unequivocal meaning 'corresponding' to the vari- ous entities it 'represents' in a more or less straightforward mirroring." I don't know how I gave Miller the impression that my "theory of lan-

guage is implicitly mimetic," a "straightforward mirror" of the reality it reflects,3 except on the assumption he seems to share with Derrida, and which seems to me obviously mistaken, that all views of language which are not in the deconstructive mode are mimetic views. My view of lan-

guage, as it happens, is by and large functional and pragmatic: language, whether spoken or written, is the use of a great variety of speech-acts to

accomplish a great diversity of human purposes; only one of these many purposes is to assert something about a state of affairs; and such a

linguistic assertion does not mirror, but serves to direct attention to selected aspects of that state of affairs.

At any rate, I think it is quite true that many of the passages I cite are equivocal and multiplex in meaning. All I claim-all that any traditional historian needs to claim-is that, whatever else the author also meant, he meant, at a sufficient approximation, at least this, and that the "this" that I specify is sufficient to the story I undertake to tell. Other historians, having chosen to tell a different story, may in their interpretation identify different aspects of the meanings conveyed by the same

passage. That brings me to the crux of my disagreement with Hillis Miller.

His central contention is not simply that I am sometimes, or always, wrong in my interpretation, but instead that I-like other traditional historians-can never be right in my interpretation. For Miller assents to Nietzsche's challenge of "the concept of 'rightness' in interpretation," and to Nietzsche's assertion that "the same text authorizes innumerable

interpretations (Auslegungen): there is no 'correct' interpretation."4 Nietzsche's views of interpretation, as Miller says, are relevant to the recent deconstructive theorists, including Jacques Derrida and himself, who have "reinterpreted Nietzsche" or have written "directly or indi-

rectly under his aegis." He goes on to quote a number of statements from Nietzsche's The Will to Power to the effect, as Miller puts it, "that

reading is never the objective identifying of a sense but the importation

3. Ibid., pp. 10-11. 4. Ibid., pp. 8, 12.

Critical Inquiry


of meaning into a text which has no meaning 'in itself.'" For example: "Ultimately, man finds in things nothing but what he himself has im-

ported into them." "In fact interpretation is itself a means of becoming master of something."5 On the face of it, such sweeping deconstructive claims might suggest those of Lewis Carroll's linguistic philosopher, who asserted that meaning is imported into a text by the interpreter's will to

power:

"The question is," said Alice, "whether you can make words mean so many different things."

"The question is," said Humpty Dumpty, "which is to be master-that's all."

But of course I don't at all believe that such deconstructive claims are, in Humpty Dumpty fashion, simply dogmatic assertions. Instead, they are conclusions which are derived from particular linguistic premises. I want, in the time remaining, to present what I make out to be the elected linguistic premises, first of Jacques Derrida, then of Hillis Miller, in the confidence that if I misinterpret these theories, my errors will soon be challenged and corrected. Let me eliminate suspense by saying at the beginning that I don't think that their radically skeptical conclusions from these premises are wrong. On the contrary, I believe that their conclusions are right-in fact, they are infallibly right, and that's where the trouble lies.

1

It is often said that Derrida and those who follow his lead subordi- nate all inquiries to a prior inquiry into language. This is true enough, but not specific enough, for it does not distinguish Derrida's work from what Richard Rorty calls "the linguistic turn"6 which characterizes modern Anglo-American philosophy and also a great part of Anglo- American literary criticism, including the "New Criticism," of the last

half-century. What is distinctive about Derrida is first that, like other French structuralists, he shifts his inquiry from language to ecriture, the written or printed text; and second that he conceives a text in an ex-

traordinarily limited fashion. Derrida's initial and decisive strategy is to disestablish the priority,

in traditional views of language, of speech over writing. By priority I mean the use of oral discourse as the conceptual model from which to derive the semantic and other features of written language and of lan-

guage in general. And Derrida's shift of elementary reference is to a

5. Ibid. 6. Richard Rorty, ed., The Linguistic Turn (Chicago and London, 1967).

428 M. H. Abrams

Spring 1977 429

written text which consists of what we find when we look at it-to "un texte deja ecrit, noir sur blanc."7 In the dazzling play of Derrida's exposi- tions, his ultimate recourse is to these black marks on white paper as the sole things that are actually present in reading, and so are not fictitious constructs, illusions, phantasms; the visual features of these black-on- blanks he expands in multiple dimensions of elaborately figurative significance, only to contract them again, at telling moments, to their elemental status. The only things that are patently there when we look at a text are "marks" that are demarcated, and separated into groups, by "blanks"; there are also "spaces," "margins," and the "repetitions" and "differences" that we find when we compare individual marks and groups of marks. By his rhetorical mastery Derrida solicits us to follow him in his move to these new premises, and to allow ourselves to be locked into them. This move is from what he calls the closed "logocen- tric" model of all traditional or "classical" views of language (which, he maintains, is based on the illusion of a Platonic or Christian transcendent being or presence, serving as the origin and guarantor of meanings) to what I shall call his own graphocentric model, in which the sole presences are marks-on-blanks.

By this bold move Derrida puts out of play, before the game even begins, every source of norms, controls, or indicators which, in the ordi-

nary use and experience of language, set a limit to what we can mean and what we can be understood to mean. Since the only givens are

already-existing marks, "deja ecrit," we are denied recourse to a speaking or writing subject, or ego, or cogito, or consciousness, and so to any possible agency for the intention of meaning something ("vouloir dire"); all such agencies are relegated to the status of fictions generated by language, readily dissolved by deconstructive analysis. By this move he leaves us no place for referring to how we learn to speak, understand, or read language, and how, by interaction with more competent users and by our own developing experience with language, we come to recognize and correct our mistakes in speaking or understanding. The author is translated by Derrida (when he's not speaking in the momentary short- hand of traditional fictions) to a status as one more mark among other marks, placed at the head or the end of a text or set of texts, which are denominated as "bodies of work identified according to the 'proper name' of a signature."8 Even syntax, the organization of words into a significant sentence, is given no role in determining the meanings of component words, for according to the graphocentric model, when we look at a page we see no organization but only a "chain" of grouped marks, a sequence of individual signs.

It is the notion of"the sign" that allows Derrida a limited opening- 7. Jacques Derrida, "La Double seance," in La Dissemination (Paris, 1972), p. 203. 8. Derrida, "La Mythologie blanche: la metaphore dans le texte philosophique," in

Marges de la philosophie (Paris, 1972), p. 304. Translations throughout are my own.

Critical Inquiry

430 M. H. Abrams The Deconstructive Angel

out of his premises. For he brings to a text the knowledge that the marks on a page are not random markings, but signs, and that a sign has a dual

aspect as signifier and signified, signal and concept, or mark-with-

meaning. But these meanings, when we look at a page, are not there, either as physical or mental presences. To account for significance, Der- rida turns to a highly specialized and elaborated use of Saussure's notion that the identity either of the sound or of the signification of a sign does not consist in a positive attribute, but in a negative (or relational) attribute-that is, its "difference," or differentiability, from other sounds and other significations within a particular linguistic system.9 This notion of difference is readily available to Derrida, because inspec- tion of the printed page shows that some marks and sets of marks repeat each other, but that others differ from each other. In Derrida's theory "difference"-not "the difference between a and b and c . .." but simply "difference" in itself-supplements the static elements of a text with an essential operative term, and as such (somewhat in the fashion of the term "negativity" in the dialectic of Hegel) it performs prodigies. For "difference" puts into motion the incessant play (jeu) of signification that goes on within the seeming immobility of the marks on the printed page.

To account for what is distinctive in the signification of a sign, Derrida puts forward the term "trace," which he says is not a presence, though it functions as a kind of"simulacrum" of a signified presence. Any signification that difference has activated in a signifier in the past re- mains active as a "trace" in the present instance as it will in the future,10 and the "sedimentation" of traces which a signifier has accumulated constitutes the diversity in the play of its present significations. This trace is an elusive aspect of a text which is not, yet functions as though it were; it plays a role without being "present"; it "appears/disappears"; "in

presenting itself it effaces itself."'1 Any attempt to define or interpret the

significance of a sign or chain of signs consists in nothing more than the

interpreter's putting in its place another sign or chain of signs, "sign- substitutions," whose self-effacing traces merely defer laterally, from substitution to substitution, the fixed and present meaning (or the

signified "presence") we vainly pursue. The promise that the trace seems to offer of a presence on which the play of signification can come to rest in a determinate reference is thus never realizable, but incessantly deferred, put off, delayed. Derrida coins what in French is the portmanteau term differance (spelled -ance, and fusing the notions of differing and

deferring) to indicate the endless play of generated significances, in

9. Ferdinand de Saussure, Course in General Linguistics, trans. Wade Baskin (New York, 1959), pp. 117-21.

10. Derrida, "La Differance," in Marges de la philosophie, pp. 12-14, 25. 11. Ibid., pp. 23-24.

Spring 1977 431

which the reference is interminably postponed.'2 The conclusion, as Derrida puts it, is that "the central signified, the originating or transcen- dental signified" is revealed to be "never absolutely present outside a system of differences," and this "absence of an ultimate signified extends the domain and play of signification to infinity."'3

What Derrida's conclusion comes to is that no sign or chain of signs can have a determinate meaning. But it seems to me that Derrida reaches this conclusion by a process which, in its own way, is no less

dependent on an origin, ground, and end, and which is no less remorse- lessly "teleological," than the most rigorous of the metaphysical systems that he uses his conclusions to deconstruct. His origin and ground are his graphocentric premises, the closed chamber of texts for which he invites us to abandon our ordinary realm of experience in speaking, hearing, reading, and understanding language. And from such a beginning we move to a foregone conclusion. For Derrida's chamber of texts is a sealed echo-chamber in which meanings are reduced to a ceaseless echolalia, a vertical and lateral reverberation from sign to sign of ghostly non-presences emanating from no voice, intended by no one, referring to nothing, bombinating in a void.

For the mirage of traditional interpretation, which vainly undertakes to determine what an author meant, Derrida proposes the alterna- tive that we deliver ourselves over to a free participation in the infinite free-play of signification opened out by the signs in a text. And on this cheerless prospect of language and the cultural enterprise in ruins Der- rida bids us to try to gaze, not with a Rousseauistic notalgia for a lost security as to meaning which we never in fact possessed, but instead with "a Nietzschean affirmation, the joyous affirmation of the play of the world and of the innocence of becoming, the affirmation of a world of signs without error [faute], without truth, without origin, which is offered to an active interpretation.. . . And it plays without security.... In absolute chance, affirmation also surrenders itself to genetic indeterminacy, to the seminal chanciness [aventure] of the trace."'4 The graphocentric premises eventuate in what is patently a metaphysics, a world-view of the free and unceasing play of differance which (since we can only glimpse this world by striking free of language, which inescapably implicates the entire

12. In the traditional or "classical" theory of signs, as Derrida describes the view that he dismantles, the sign is taken to be "a deferred presence . . . the circulation of signs defers the moment in which we will be able to encounter the thing itself, to get hold of it, consume or expend it, touch it, see it, have a present intuition of it" (ibid., p. 9). See also "Hors livre" in La Dissemination, pp. 10-11.

13. Derrida, "La Structure, le signe et lejeu dans le discours des sciences humaines," in L'Ecriture et la difference (Paris, 1967), p. 411.

14. Ibid., p. 427. Derrida adds that this "interpretation of interpretation," which "affirms free-play ... tries to pass beyond man and humanism...." On the coming "monstrosity," see also De la grammatologie (Paris, 1967), p. 14.

Critical Inquiry


metaphysics of presence that this view replaces) we are not able even to name. Derrida's vision is thus, as he puts it, of an "as yet unnamable something which cannot announce itself except ... under the species of a non-species, under the formless form, mute, infant, and terrifying, of monstrosity."15

2

Hillis Miller sets up an apt distinction between two classes of current structuralist critics, the "canny critics" and the "uncanny critics." The canny critics cling still to the possibility of"a structuralist-inspired criticism as a rational and rationalizable activity, with agreed-upon rules of procedure, given facts, and measurable results." The uncanny critics have renounced such a nostalgia for impossible certainties.'6 And as himself an uncanny critic, Miller's persistent enterprise is to get us to share, in each of the diverse works that he criticizes, its self- deconstructive revelation that in default of any possible origin, ground, presence, or end, it is an interminable free-play of indeterminable mean-

ings. Like Derrida, Miller sets up as his given the written text, "innocent

black marks on a page"17 which are endowed with traces, or vestiges of

meaning; he then employs a variety of strategies that maximize the number and diversity of the possible meanings while minimizing any factors that might limit their free-play. It is worthwhile to note briefly two of those strategies.

For one thing Miller applies the terms "interpretation" and "mean-

ing" in an extremely capacious way, so as to conflate linguistic utterance or writing with any metaphysical representation of theory or of "fact" about the physical world. These diverse realms are treated equivalently as "texts" which are "read" or "interpreted." He thus leaves no room for

taking into account that language, unlike the physical world, is a cultural institution that developed expressly in order to mean something and to convey what is meant to members of a community who have learned how to use and interpret language. And within the realm of explicitly verbal texts, Miller allows for no distinction with regard to the kinds of norms that may obtain or may not obtain for the "interpretation" of the entire

15. Derrida, "La Structure, le signe," p. 428. "We possess no language . . . which is alien to this history; we cannot express a single destructive proposition which will not

already have slipped into the form, the logic, and the implicit postulates of that very thing that it seeks to oppose." "Each limited borrowing drags along with it all of metaphysics" (pp. 412-13).

16. J. Hillis Miller, "Stevens' Rock and Criticism as Cure, II," The Georgia Review 30

(Summer 1976): 335-36. 17. Miller, "Walter Pater: A Partial Portrait," Daedalus 105 (Winter 1976): 107.

432 M. H. Abrams

Spring 1977 433

corpus of an individual author's writings, or of a single work in its totality, or of a particular passage, sentence, or word within that work. As a critical pluralist, I would agree that there are a diversity of sound (though not equally adequate) interpretations of the play King Lear, yet I claim to know precisely what Lear meant when he said, "Pray you undo this button."

A second strategy is related to Derrida's treatment of the "trace." Like Derrida, Miller excludes by his elected premises any control or limitation of signification by reference to the uses of a word or phrase that are current at the time an author writes, or to an author's intention, or to the verbal or generic context in which a word occurs. Any word within a given text-or at least any "key word," as he calls it, that he picks out for special scrutiny-can thus be claimed to signify any and all of the diverse things it has signified in the varied forms that the signifier has assumed through its recorded history; and not only in a particular lan-

guage, such as English or French, but back through its etymology in Latin and Greek all the way to its postulated Indo-European root. Whenever and by whomever and in whatever context a printed word is used, therefore, the limits of what it can be said to mean in that use are set only by what the interpreter can find in historical and etymological dictionaries, supplemented by any further information that the

interpreter's own erudition can provide. Hence Miller's persistent recourse to etymology-and even to the significance of the shapes of the printed letters in the altering form of a word-in expounding the texts to which he turns his critical attention.18

Endowed thus with the sedimented meanings accumulated over its total history, but stripped of any norms for selecting some of these and

rejecting others, a key word-like the larger passage or total text of which the word is an element-becomes (in the phrase Miller cites from Mallarme) a suspens vibratoire,19 a vibratory suspension of equally likely meanings, and these are bound to include "incompatible" or "irreconcil- able" or "contradictory" meanings. The conclusion from these views Miller formulates in a variety of ways: a key word, or a passage, or a text, since it is a ceaseless play of anomalous meanings, is "indeterminable," "undecipherable," "unreadable," "undecidable."20 Or more bluntly: "All reading is misreading." "Any reading can be shown to be a misreading on evidence drawn from the text itself." But in misreading a text, the

interpreter is merely repeating what the text itself has done before him,

18. See, for example, his unfolding of the meanings of "cure" and "absurd" in "Ste- vens' Rock and Criticism as Cure," I, The Georgia Review 30 (Spring 1976): 6-11. For his

analysis of significance in the altering shapes, through history, of the printed form of a word see his exposition of abyme, ibid., p. 11; also his exposition of the letter x in "Ariadne's Thread: Repetition and the Narrative Line," Critical Inquiry 3 (Autumn 1976): 75-76.

19. "Tradition and Difference," p. 12. 20. See, e.g., "Stevens' Rock," I, pp. 9-11; "Walter Pater," p. 111.

Critical Inquiry


for "any literary text, with more or less explicitness or clarity, already reads or misreads itself."21 To say that this concept of interpretation cuts the ground out from under the kind of history I undertook to write is to take a very parochial view of what is involved; for what it comes to is that no text, in part or whole, can mean anything in particular, and that we can never say just what anyone means by anything he writes.

But if all interpretation is misinterpretation, and if all criticism (like all history) of texts can engage only with a critic's own misconstruction, why bother to carry on the activities of interpretation and criticism? Hillis Miller poses this question more than once. He presents his answers in terms of his favorite analogues for the interpretive activity, which he explores with an unflagging resourcefulness. These analogues figure the text we read as a Cretan labyrinth, and also as the texture of a spider's web; the two figures, he points out, have been fused in earlier conflations in the myth of Ariadne's thread, by which Theseus retraces the windings of the labyrinth, and of Arachne's thread, with which she spins her web.22 Here is one of Miller's answers to the question, why pursue the critical enterprise?

Pater's writings, like those of other major authors in the Occidental tradition, are at once open to interpretation and ultimately inde- cipherable, unreadable. His texts lead the critic deeper and deeper into a labyrinth until he confronts a final aporia. This does not mean, however, that the reader must give up from the beginning the attempt to understand Pater. Only by going all the way into the labyrinth, following the thread of a given clue, can the critic reach the blind alley, vacant of any Minotaur, that impasse which is the end point of interpretation.23

Now, I make bold to claim that I understand Miller's passage, and that what it says, in part, is that the deconstructive critic's act of interpretation has a beginning and an end; that it begins as an intentional, goal- oriented quest; and that this quest is to end in an impasse.

The reaching of the interpretive aporia or impasse precipitates what Miller calls "the uncanny moment"-the moment in which the critic, thinking to deconstruct the text, finds that he has simply participated in the ceaseless play of the text as a self-deconstructive artefact. Here is another of Miller's statements, in which he describes both his own and Derrida's procedure:

Deconstruction as a mode of interpretation works by a careful and circumspect entering of each textual labyrinth. . . . The deconstructive critic seeks to find, by this process of retracing, the ele-

21. "Walter Pater," p. 98; "Stevens' Rock, II," p. 333. 22. "Ariadne's Thread," p. 66. 23. "Walter Pater," p. 112.

434 M. H. Abrams

Spring 1977 435

ment in the system studied which is alogical, the thread in the text in question which will unravel it all, or the loose stone which will pull down the whole building. The deconstruction, rather, annihi- lates the ground on which the building stands by showing that the text has already annihilated that ground, knowingly or unknow- ingly. Deconstruction is not a dismantling of the structure of a text but a demonstration that it has already dismantled itself.24

The uncanny moment in interpretation, as Miller phrases it elsewhere, is a sudden "mise en abyme" in which the bottom drops away and, in the endless regress of the self-baffling free-play of meanings in the very signs which both reveal an abyss and, by naming it, cover it over, we catch a glimpse of the abyss itself in a "vertigo of the underlying nothingness."25

The "deconstructive critic," Miller has said, "seeks to find" the alogical element in a text, the thread which, when pulled, will unravel the whole texture. Given the game Miller has set up, with its graphocentric premises and freedom of interpretive maneuver, the infallible rule of the deconstructive quest is, "Seek and ye shall find." The deconstructive method works, because it can't help working; it is a can't-fail enterprise; there is no complex passage of verse or prose which could possibly serve as a counter-instance to test its validity or limits. And the uncanny critic, whatever the variousness and distinctiveness of the texts to which he

applies his strategies, is bound to find that they all reduce to one thing and one thing only. In Miller's own words: each deconstructive reading, "performed on any literary, philosophical, or critical text ... reaches, in the particular way the given text allows it, the 'same' moment of an aporia. . . . The reading comes back again and again, with different texts, to the 'same' impasse."26

It is of no avail to point out that such criticism has nothing whatever to do with our common experience of the uniqueness, the rich variety, and the passionate human concerns in works of literature, philosophy, or criticism-these are matters which are among the linguistic illusions that the criticism dismantles. There are, I want to emphasize, rich rewards in reading Miller, as in reading Derrida, which include a delight in his resourceful play of mind and language and the many and striking insights yielded by his wide reading and by his sharp eye for unsus- pected congruities and differences in our heritage of literary and philosophical writings. But these rewards are yielded by the way, and that way is always to the ultimate experience of vertigo, the uncanny

24. "Stevens' Rock, II," p. 341. See also "Walter Pater," p. 101, and "Ariadne's Thread," p. 74.

25. "Stevens' Rock," I, pp. 11-12. The unnamable abyss which Miller glimpses has its

parallel in the unnamable and terrifying monstrosity which Derrida glimpses; see above, p. 432.

26. "Deconstructing the Deconstructors," Diacritics 5 (Summer 1975): 30.

Critical Inquiry

436 M. H. Abrams The Deconstructive Angel

frisson at teetering with him on the brink of the abyss; and even the shock of this discovery is soon dulled by its expected and invariable recurrence.

I shall cite a final passage to exemplify the deft and inventive play of Miller's rhetoric, punning, and figuration, which give his formulations of the mise en abyme a charm that is hard to resist. In it he imposes his fused analogues of labyrinth and web and abyss on the black-on-blanks which constitute the elemental given of the deconstructive premises:

Far from providing a benign escape from the maze, Ariadne's thread makes the labyrinth, is the labyrinth. The interpretation or solving of the puzzles of the textual web only adds more filaments to the web. One can never escape from the labyrinth because the activity of escaping makes more labyrinth, the thread of a linear narrative or story. Criticism is the production of more thread to embroider the texture or textile already there. This thread is like a filament of ink which flows from the pen of the writer, keeping him in the web but suspending him also over the chasm, the blank page that thin line hides.27

To interpret: Hillis Miller, suspended by the labyrinthine lines of a textual web over the abyss that those black lines demarcate on the blank

page, busies himself to unravel the web that keeps him from plunging into the blank-abyss, but finds he can do so only by an act of writing which spins a further web of lines, equally vulnerable to deconstruction, but only by another movement of the pen that will trace still another

inky net over the ever-receding abyss. As Miller remarks, I suppose ruefully, at the end of the passage I quoted, "In one version of Ariadne's

story she is said to have hanged herself with her thread in despair after

being abandoned by Theseus."

3

What is one to say in response to this abysmal vision of the textual world of literature, philosophy, and all the other achievements of man- kind in the medium of language? There is, I think, only one adequate response, and that is the one that William Blake made to the Angel in The Marriage of Heaven and Hell. After they had groped their way down a

"winding cavern," the Angel revealed to Blake a ghastly vision of hell as an "infinite Abyss"; in it was "the sun, black but shining," around which were "fiery tracks on which revolv'd vast spiders." But no sooner, says Blake, had "my friend the Angel" departed, "then this appearance was no more, but I found myself sitting on a pleasant bank beside a river by moon light, hearing a harper who sung to a harp." The Angel, "sur-

27. "Stevens' Rock, II," p. 337.

Critical Inquiry Spring 1977

prised asked me how I escaped? I answered: 'All that we saw was owing to your metaphysics.' "

As a deconstructive Angel, Hillis Miller, I am happy to say, is not serious about deconstruction, in Hegel's sense of "serious"; that is, he does not entirely and consistently commit himself to the consequences of his premises. He is in fact, fortunately for us, a double agent who plays the game of language by two very different sets of rules. One of the games he plays is that of a deconstructive critic of literary texts. The other is the game he will play in a minute or two when he steps out of his graphocentric premises onto this platform and begins to talk to us.

I shall hazard a prediction as to what Miller will do then. He will have determinate things to say and will masterfully exploit the resources of language to express these things clearly and forcibly, addressing himself to us in the confidence that we, to the degree that we have mastered the constitutive norms of this kind of discourse, will approximate what he means. He will show no inordinate theoretical difficulties about beginning his discourse or conducting it through its middle to an end. What he says will manifest, by immediate inference, a thinking subject or ego and a distinctive and continuant ethos, so that those of you who, like myself, know and admire his recent writings will be surprised and de- lighted by particularities of what he says, but will correctly anticipate both its general tenor and its highly distinctive style and manner of proceeding. What he says, furthermore, will manifest a feeling as well as thinking subject; and unless it possesses a superhuman forbearance, this subject will express some natural irritation that I, an old friend, should so obtusely have misinterpreted what he has said in print about his critical intentions.

Before coming here, Miller worked his thoughts (which involved inner speech) into the form of writing. On this platform, he will proceed to convert this writing to speech; and it is safe to say-since our chairman is himself a double agent, editor of a critical journal as well as

organizer of this symposium-that soon his speech will be reconverted to writing and presented to the public. This substitution of ecriture for parole will certainly make a difference, but not an absolute difference; what Miller says here, that is, will not jump an ontological gap to the printed page, shedding on the way all the features that made it intelligi- ble as discourse. For each of his readers will be able to reconvert the black-on-blanks back into speech, which he will hear in his mind's ear; he will perceive the words not simply as marks nor as sounds, but as already invested with meaning; also, by immediate inference, he will be aware in his reading of an intelligent subject, very similar to the one we will infer while listening to him here, who organizes the well-formed and significant sentences and marshals the argument conveyed by the text.

There is no linguistic or any other law we can appeal to that will prevent a deconstructive critic from bringing his graphocentric proce-

437


dures to bear on the printed version of Hillis Miller's discourse-or of mine, or of Wayne Booth's-and if he does, he will infallibly be able to translate the text into a vertiginous mise en abyme. But those of us who

stubbornly refuse to substitute the rules of the deconstructive enterprise for our ordinary skill and tact at language will find that we are able to understand this text very well. In many ways, in fact, we will understand it better than while hearing it in the mode of oral discourse, for the institution of print will render the fleeting words of his speech by a durable graphic correlate which will enable us to take our own and not the speaker's time in attending to it, as well as to re-read it, to collocate, and to ponder until we are satisfied that we have approximated the author's meaning.

After Hillis Miller and I have pondered in this way over the text of the other's discourse, we will probably, as experience in such matters indicates, continue essentially to disagree. By this I mean that neither of us is apt to find the other's reasons so compelling as to get him to change his own interpretive premises and aims. But in the process, each will have come to see more clearly what the other's reasons are for doing what he does, and no doubt come to discover that some of these reasons are indeed good reasons in that, however short of being compelling, they have a bearing on the issue in question. In brief, insofar as we set ourselves, in the old-fashioned way, to make out what the other means

by what he says, I am confident that we shall come to a better mutual

understanding. After all, without that confidence that we can use lan-

guage to say what we mean and can interpret language so as to determine what was meant, there is no rationale for the dialogue in which we are now engaged.

438 M. H. Abrams

Documents

Deconstructive angel